Skip to main content
What Went Down On Super Tuesday

Leave a comment, and send us questions @FiveThirtyEight.

12:24 AM
A Super Goodnight

That’s a wrap, everyone! It was certainly a super night for Clinton. It was semi-super for Trump. If you want to relive the Super Tuesday results chronologically, click here to start at the beginning of our live blog and scroll up.

We’ll have a lot more to say on what the Super Tuesday results mean for Cruz, Rubio, Kasich and Sanders for you to chew over with your morning croissant. In the meantime, though, here’s where things stand as we shut the lights off on this live blog:

Alabama Trump Alabama Clinton
Arkansas Trump Arkansas Clinton
Georgia Trump Massachusetts Clinton
Massachusetts Trump Georgia Clinton
Tennessee Trump Tennessee Clinton
Virginia Trump Texas Clinton
Oklahoma Cruz Virginia Clinton
Texas Cruz American Samoa Clinton
Minnesota Rubio Colorado Sanders
Alaska Not called Minnesota Sanders
Vermont Trump Oklahoma Sanders
Vermont Sanders
Projected winners so far

Source: ABC News

12:15 AM
We're Still In The Interpretation Phase

Responding to Ben’s “looking ahead” comments, a couple of things jumped out at me. He mentioned the economy, and this seems like something that’s been lost because of the recent — and important — discussions about race and immigration. I wrote after Iowa that a possible split for Democrats was between the winners and losers of the Obama years. This isn’t all that surprising for the incumbent party during a soft economic recovery. But it would certainly be interesting if this continues to distinguish Trump voters from Republicans who go for Rubio or Cruz.

So that’s really important for what the results of each contest mean. As someone who studies election interpretation for a living, I love it. But I mentioned at the beginning of this evening’s discussion, the interpretation stage usually ends now, as one candidate pulls ahead in delegates and voter support. I think we can now say that for Republicans, that still hasn’t decisively happened, even though Trump had a good night. The interpretations have already begun — we can no longer say Rubio hasn’t won anything. Cruz has beaten Trump three times. At some point, the time for interpretation will be over and the result will just be the result. Right?

12:12 AM
Pollsters React To Super Tuesday

Since we cover polls so much here and rely on them for our forecasts, we occasionally check in with the people who produce the polls released to the public — most recently, just after the Iowa caucuses. Tonight, we asked a handful of pollsters to react to the results. Lots of polls were in line with today’s results, but there were also possible polling misses in Massachusetts — understating Sanders’s support — and in Oklahoma, where polls overstated Trump’s chances. We heard back from four pollsters, who said polls mostly had a good night and that Clinton-Trump remains the most likely general-election matchup.

Oklahoma was a “big miss in the late polling,” said Christopher B. Budzisz, poll director at Loras College, in Dubuque, Iowa.

Matthew Towery of Opinion Savvy noted that Oklahoma has a closed primary, which he said seems to be the main reason for the poll-results differential. “I assume that pollsters failed to screen adequately in Oklahoma,” he said, “which would account for the apparent inflation of Trump’s numbers via independents.”

As for Massachusetts, said Spencer H. Kimball, adviser to the Emerson College Polling Society, it “looked like Bernie was able to get out the younger vote like he did in the earlier states.”

I asked the pollsters to estimate the probabilities that Trump and Clinton would win their respective parties’ nominations. On Trump, they ranged from 60 percent (Budzisz) to 90 percent (Kimball).

“He’s Teflon,” said Chris Kane, co-president of the Emerson College Polling Society. “It’s hard to see what might derail him.”

And no one gave Clinton less than a 90 percent chance of winning the Democratic nomination. Kimball thinks it’s inevitable: He said there’s a 100 percent chance Clinton will be the nominee.

12:06 AM
Back-Of-The-Envelope Delegate Projections: Tonight Sucks For Rubio

Here are my back-of-the-envelope delegate projections: Based on current vote counts and each state and district’s allocation rules, Trump is on track to end up with about 262 delegates, Cruz is on track to win about 215, and Rubio is on track to win about 93. That’s an even more awful result for Rubio than I thought, and it’s a lot worse for Rubio than The New York Times’s Upshot model is currently projecting. After tonight, Rubio could be more than 100 delegates behind Cruz and more than 200 delegates behind Trump.

11:59 PM

With the last few results still trickling in, what can tonight teach us about upcoming contests? This weekend, Republicans will vote in Louisiana, Maine and Kentucky (and also Puerto Rico), while Democrats will vote in Louisiana, Maine and Nebraska.

Economically, Louisiana and Kentucky look a lot like the poorer Southern states that voted today, all of which were comfortable wins for Trump and Clinton. Maine is a Northeastern state but looks quite different economically from the rest of New England: Its population is poorer, less educated and significantly older than its neighbors. Still, its demographics probably line up well for Sanders.

Nebraska, meanwhile, has one of the strongest economies in the country; its unemployment rate in December was just 3 percent, meaning pretty much anyone who wants a job can get one. Of the states that have voted so far, it probably most resembles Iowa economically.

Looking ahead to next week, we get primaries or caucuses in Idaho (another rural state with a relatively low unemployment rate), Mississippi (another poor Southern state with a large black population) and Hawaii (something of an outlier economically and in most other ways). We also get primaries in Michigan, which was hit harder by the recession than nearly any other state. Michigan will be the first glimpse we get of how the industrial Midwest responds to the candidates, which makes it especially worth watching.

UPDATE (March 2, 12:30 a.m.): A reader writes in (thanks, Jason!) to note that this post originally left out one state: Kansas, which also caucuses Saturday. (Was this omission due to the fact I’m a Mizzou fan? You be the judge.) Kansas is in many ways pretty representative of the U.S. economy as a whole: The state’s unemployment rate, at 4 percent, is better than the national mark of 4.9 percent, but in terms of income, poverty and education, Kansas is right around the national average. It’s more rural than the U.S. as a whole, and more of its economy comes from agriculture, but it also has a robust healthcare sector. In terms of states that have voted so far, it’s probably most similar to Oklahoma economically, with the (very large) caveat that it doesn’t have much of an energy industry.

11:58 PM
Still The Party Of Goldwater?

Super Tuesday, as I noted earlier, began as a coordinated Southern primary. We’ve seen Trump do well in New Hampshire and Massachusetts. Nate pointed out last week that this is where Trump is strong — the South and Northeast. It’s also where two of the race’s strongest contenders, Cruz and Trump, are from. (And Rubio, too, sort of. Florida is weird in the geography game.)

As I look at this list of contenders and of states that have held contests so far, one thing that strikes me is that this suddenly no longer looks like the party of Nixon, Goldwater and Reagan — the party of the West. California is pretty much a lost cause for Republicans. And yet the interior west and several of the Plains states — the Dakotas, Wyoming, Idaho, Utah, Kansas — are incredibly strong for Republican presidential candidates. Some of those states vote fairly late, but it seems possible now that the contest could go on that long (or longer). With prominent Nebraska Republican Ben Sasse leading the #NeverTrump movement, I’ll be curious to see what happens as these states head to the polls.

11:54 PM

Back in January, Trump marveled at all his free media time: “I get so much, they call it free. … They do 15 minutes on Trump and then they say, ‘Now we will be back after commercial break with something else on Trump’ … because I get good ratings.” The data back him up: He consistently has gotten the vast majority of media coverage, not just a plurality, in a race that had a dozen candidates at the time of his comment. His stump speech tonight, followed by a news conference and carried live on several networks, was the latest evidence that Trump has mastered getting free media time even as his rivals struggle to raise money for paid time.

Trump dominates both news and search
11:38 PM
Alabama Trump Alabama Clinton
Arkansas Trump Arkansas Clinton
Georgia Trump Massachusetts Clinton
Massachusetts Trump Georgia Clinton
Tennessee Trump Tennessee Clinton
Virginia Trump Texas Clinton
Oklahoma Cruz Virginia Clinton
Texas Cruz American Samoa Clinton
Minnesota Rubio Colorado Sanders
Alaska Not called Minnesota Sanders
Vermont Not called Oklahoma Sanders
Vermont Sanders
Projected winners so far

Source: ABC

11:36 PM

The betting markets are pretty consistent with what we’ve been saying about the Republican race: Rubio’s chance of becoming the nominee has fallen to 9 percent after returns started coming in, according to Predictwise. Cruz has risen to 5 percent, and Trump remains the heavy favorite, at 83 percent.

11:34 PM

As Nate noted, the timing of Rubio’s first state victory, in Minnesota, was poor, coming late in the night after Trump and Cruz each won multiple states. On the other hand, Trump’s timing of his speech-press conference was terrific. He bragged that even where he hasn’t won, he’s finished second. But Trump is coming in third in Minnesota, trailing Cruz, who is in second, by 7 percentage points. Then again, Minnesota was one of Rubio’s best states and the only one in which he won a state poll conducted in 2016 and compiled by HuffPost Pollster.

11:27 PM

The Associated Press has called the Minnesota Republican caucuses for Rubio, giving him his first win of the campaign. According to the Minnesota Republican Party, Rubio has 36 percent of the vote to Cruz’s 29 percent and Trump’s 22 percent, with 70 percent of precincts reporting.

In theory, the win would partially change the media narrative about Rubio’s evening, particularly given that Cruz had emphasized how he was the only non-Trump candidate to have won a state so far. But the call came late in the evening after the conventional wisdom had already jelled around it being a poor night for Rubio. Some of this is silly, as is the sarcastic response that political Twitter has taken toward Rubio lately. Ultimately, however, Rubio won’t shut up his doubters unless he wins his home state of Florida, a winner-take-all state that votes on March 15. Rubio currently trails Trump in polls there.

11:24 PM

Harry, my first reaction to that comment was that the difference in context is telling. In 1992, you had an incumbent who had initially looked pretty strong. It’s not surprising that this would draw lots of potentially factional Democratic candidates.

In the last two open-seat presidential primaries, which were pretty competitive (and I think both parties thought they would be), the parties were much more organized. Both parties’ primaries in 2000 almost had no contests at all, and 2008 had an exciting contest between two mainstream, strong candidates on the Democratic side.

For a party to face an open contest, with the other party the incumbent party — so basically the best possible opportunity — and be so fractured is kind of a different beast.

11:23 PM

This was interesting, from commenter Andy Hicken:

The last time 4 candidates from one party all won a state on the same night: March 3, 1992. Paul Tsongas won Maryland, Utah, and Washington, Tom Harkin won Idaho and Minnesota, Jerry Brown won Colorado, and Bill Clinton won Georgia.

It looks like it’ll only be three GOP winners tonight, unless Kasich can overcome Trump in Vermont. Still, it speaks to this point: I don’t know why folks are walking away from tonight with the idea that Trump is somehow dominant. He’s won only 37 percent of the vote, pooling across states. That means he’s ahead but far from being a lock of any sort.

11:22 PM
Forecast Models Could Be Underestimating Cruz's Delegate Margin Over Rubio

The New York Times’ Upshot tool, which projects each candidate’s Super Tuesday delegate tally, may be underestimating Cruz’s eventual delegate margin over Rubio. Right now, it forecasts that Cruz will win about 50 to 60 more delegates than Rubio. I think that may be low-balling the magnitude of Rubio’s challenge after this evening.

Think about Texas for a second. If Cruz is winning most congressional districts and Rubio falls short of the state’s 20 percent threshold, Cruz will likely win about 90 more delegates than Rubio in Texas alone, thanks to its district allocation rules. But then, consider that Cruz is also out-polling Rubio for second place in a vast majority of other congressional districts in the South. That means Trump will be winning two delegates from most Southern districts, while Cruz will be winning one, and Rubio will be winning zero.

Altogether, that could push Cruz’s differential over Rubio well above 100 delegates, perhaps close to 125. And Rubio’s strong performances in Minnesota and Virginia may not be enough to offset his failure to hit thresholds in states like Texas, Tennessee and Alabama, if current vote returns hold. If Rubio is that far behind Cruz in the count, it would bode extremely poorly for his chances of maintaining his status as the leading alternative to Trump.

11:14 PM

11:12 PM

“The only way to beat Donald Trump is to stand together united,” Cruz said tonight. But what if the only way to beat Trump was to make the No. 2 or No. 3 candidate a Hunger Games-style tribute? A piece by The Washington Post’s Alexandra Petri included a graphic that went viral … a GOP version of ethical game theory puzzle “the prisoners’ dilemma.” It’s much more plausible to me that if neither Cruz nor Rubio drops out (and neither seem primed to), then Trump has a clear path to the nomination.

11:08 PM

A follow-up to Carl’s post below about Jim Gilmore’s “victory”:

11:07 PM
Delegates: Who's On Track For The GOP Nomination?

Heading into tonight, according to FiveThirtyEight’s delegate interactive, Trump needed to hit a cumulative total of 297 delegates (including the 82 delegates he had already won in Iowa, New Hampshire, South Carolina and Nevada) to stay “on track” for the 1,237 required to clinch the nomination, while Cruz needed 384 and Rubio needed 242. Right now, Trump looks as if he’ll win somewhere between 225 and 300 delegates, easily keeping him “on pace.”

But it will be a close call whether Cruz or Rubio reaches a higher share of his delegate target. Thanks to Texas and Oklahoma, plus strong second-place finishes in many Southern congressional districts, Cruz looks likely to win as many as 100 or more delegates than Rubio. The bar for Rubio was always lower for the “SEC Primary,” but a delegate gap that wide could cause Cruz to knock Rubio out of second place on our leaderboard. It may take days to determine the final counts because in many cases, results will need to be broken down by congressional districts.

11:06 PM

NBC News just called the Colorado Democratic caucuses for Sanders, giving him his third win of the evening, along with Vermont and Oklahoma. Sanders will also probably win the Minnesota caucus later on tonight. Still, it appears that he’ll narrowly lose in Massachusetts, a state that was something of a “must-win” for him based on our demographic targets. As I wrote earlier tonight, however, the big problem for Sanders is the huge margins that Clinton is racking up in the non-Oklahoma Southern states. Since the Democrats’ delegate allocation is highly proportional, that’s a deficit he’ll have to make up later on.

11:00 PM

The biggest news in Oklahoma tonight might not be Cruz’s victory, or Sanders’s, but rather the indictment of Aubrey K. McClendon on antitrust charges. McClendon may not be a household name nationally, but he’s a genuine celebrity in Oklahoma, where he founded natural gas giant Chesapeake Energy Corp. and built it into one of the biggest oil and gas producers in the U.S. before being pushed out in 2013 amid a shareholder revolt. In an industry that has largely left its wildcatter days behind, McClendon was one of the last truly giant personalities — he was charismatic, brilliant and reckless. After losing a fortune to a margin call in 2008, he told me that it was a “kick in the shins” but that he was more bothered by his Oklahoma Sooners losing to Colt McCoy’s Texas Longhorns. It was braggadocio, but I half-believed it.

10:57 PM

Micah, it’s true that Clinton overwhelmingly won the black vote in Super Tuesday states where she beat Sanders, including Georgia. Pragmatism about black political interests and how the game is played is likely the primary factor, since Sanders has also spoken to issues of core interest to black voters.

But a candidate speaking to the issues that a demographic cares about isn’t enough, no matter your race, and particularly so for black voters. Many black voters could support Sanders’s positions, but if they don’t think he knows how to wrangle Congress, there’s a risk in voting for him. I can’t help but think of President Lyndon B. Johnson wrangling an ambivalent Congress to pass civil rights legislation. He was known for his ability to work inside the political system, which may be tactically more important for black voters than white voters.

I’ve seen some self-described white Sanders voters express anger on social media, saying that black people are voting against their interests. But one of the roles the president plays is interacting with Congress and pushing (or aiming to block) the passage of legislation. And black and white voters have very different experiences with government when it comes to supporting legislation. This University of Chicago study shows how, all other factors aside, black support for legislation means it’s less likely to be passed.If white voters support a bill, it’s much more likely to be passed and adopted. But if black voters support legislation, it’s actually less likely to pass. That argues that black voters may have a tactical interest in an establishment candidate they think can work behind the scenes in their interest, and there’s a perception that Clinton may be better at insider politics. That also tracks with the broader support on the Democratic side for an experienced candidate, versus on the GOP side for an anti-establishment candidate.

10:57 PM

Farai, so after tonight’s results, it’s probably safe to say black voters just aren’t buying what Sanders is selling. Why? Are there a lot of causes? One big one?

10:54 PM
Gender Men 42% 53%
Women 58 65
Race Whites 58 50
Blacks 25 83
Hispanics 12 63
Age 17-29 18 36
30-44 22 55
45-64 41 67
65+ 19 73
Party Democrats 70 69
Independents 26 38
Republicans 3 33
ABC analysis of exit polls for all Democratic primaries
10:54 PM

How did Jim Gilmore, the former Virginia governor who perpetually polled at 0 percent in national polls before withdrawing from the race three weeks ago, win 47 percent of the Republican primary vote in Chelsea, Massachusetts? I have no idea, but the Boston Globe results page shows Gilmore got 366 votes in Chelsea — or 42 percent of his total in the state. The Atlantic and the Guardian have reported this stunning result. Occam’s razor suggests that this could be a typo or other error — it would mean Gilmore got more than 10 times as many votes in Chelsea (est. pop. 37,000) as he got in the rest of Suffolk County (est. pop. 767,000).

10:51 PM
Is Trump A Factional Candidate?

There have been lots of attempts to place Trump in some kind of historical context — around Labor Day, I theorized that this was a function of his rogue, insurgent candidacy and our need to place him in some sort of context. Now that he seems to be winning, historically minded political scientists have compared him with a few presidential candidates who just couldn’t unite their parties: George McGovern, Barry Goldwater and William Jennings Bryan (in 1896). But each of these candidates, in different ways, represented a movement, a world view, and an ideological faction within their parties.

Trump’s candidacy is in some sense a test of that — and the emergence of a faction unified by race and immigration issues is still a live possibility. But Cruz’s speech tonight, which attempted to cast Trump as a liberal who would expand Obamacare, support Planned Parenthood and otherwise be indistinguishable from a Democrat, speaks to something else. It may well be the case that Trump will be nominated and subsequently won’t be able to unify the party around his candidacy. But despite Trump’s insistence that his campaign has been a movement, the ideological meaning of his candidacy remains murkier and less defined than that of factional candidates of the past.

10:50 PM

Sanders is not having a good night. Last week, we published demographic targets in each state. These were not predictions; instead, they were estimates, based on the racial composition of each state and other demographics, of how well a candidate would have to do to get half of pledged delegates nationwide.

Alabama Clinton +59 Clinton +30 Clinton +29
Arkansas Clinton +41 Clinton +15 Clinton +26
Colorado Sanders +6 Sanders +11 Clinton +5
Georgia Clinton +43 Clinton +27 Clinton +16
Massachusetts Clinton +3 Sanders +11 Clinton +14
Minnesota Sanders +18 Sanders +21 Clinton +3
Oklahoma Sanders +11 Sanders +4 Sanders +7
Tennessee Clinton +35 Sanders +2 Clinton +37
Texas Clinton +37 Clinton +13 Clinton +24
Vermont Sanders +72 Sanders +83 Clinton +11
Virginia Clinton +29 Clinton +9 Clinton +20
Average Clinton +13 Sanders +3 Clinton +16
Sanders underperforming demographic targets

Results as of 10:40 p.m.

Clinton is running ahead of her benchmarks by an average of 16 percentage points tonight, which is equivalent to her holding a 16-point lead over Sanders in national polls. Sanders won a few of his “must-win” states tonight, but not others, and the huge deficits he racked up to Clinton in Southern states will make it hard for him to make up his deficit later on.

10:45 PM

Here’s a bad sign for Rubio’s bid to hit the 20 percent delegate threshold in Texas: In the handful of counties I’ve checked, he’s doing WORSE with election day voters than early voters. Most votes counted so far are early ballots, and he’s at just 18 percent. This would be a heartbreaking blow to Rubio’s delegate count.

10:44 PM

Fox News called Massachusetts for Clinton. If that is correct, it’s a symbolic blow to Sanders. As I’ve mentioned over and over again, Clinton is running up huge delegate margins in the South. Sanders was hoping to win a few states despite that. And while he is likely to win four states (Colorado, Minnesota, Oklahoma and Vermont), those won’t come anywhere near helping him cut into Clinton’s large delegate lead.

10:39 PM

Earlier today, I broke the Super Tuesday states into three categories based on their economic characteristics. So far, those groupings seem to be holding up reasonably well in the actual voting, at least on the Republican side.

With the Associated Press calling Arkansas for Trump, it now looks like Trump will win all four of the states I labeled “Southern strugglers” (Alabama, Arkansas, Georgia and Tennessee), in most cases by large margins. Cruz has won both of the “energy boomers” that have reported so far (Texas and Oklahoma), with Alaska results still outstanding. The third group, wealthier states I labeled “educated elites,” are a bit more mixed: Trump won Massachusetts overwhelmingly and Virginia narrowly. In Vermont, he’s battling with Kasich (!), and he looks likely to lose Minnesota to either Rubio or Cruz.

The Democratic race isn’t aligning as neatly with my economic groupings. As I noted earlier, Clinton won all the Southern states, but she lost Oklahoma despite winning Texas. The candidates have split the “educated elite” states so far, with Massachusetts, Minnesota and Colorado still outstanding.

10:35 PM

At The New York Times, the Upshot is running constantly updated delegate estimates as the results come in, and the estimates are changing based on whether candidates are meeting qualifying delegate thresholds, among other things. (For example, Rubio is close to the 20 percent of the vote that he’d need to win statewide delegates in Texas.)

As of 10:30 p.m. Eastern time, however, they’re projecting Trump to get around 245 delegates from the voting tonight, Cruz about 180 and Rubio about 120, Kasich around 20 and Carson about 5. That would mean Trump gets around 43 percent of delegates from the evening, a good showing for him, but not quite as good as it looked based on polls entering the evening, when a majority of delegates seemed possible. Also a pretty good result for Cruz, although a lot of his strongest states voted tonight.

10:31 PM

Trump adds Arkansas to his wins tonight. No doubt, a victory is good for Trump, but the delegate totals coming out of the state will be fairly even as Cruz, Trump and Rubio are all safely above the statewide delegate threshold.

10:28 PM

10:28 PM

Clinton is barely holding on in Massachusetts. She has a 3.6 percentage point lead with 66 percent of precincts reporting. That margin will likely drop, although at this point, it doesn’t seem like it’ll be enough for Sanders to overtake her. If her advantage continues, that will mean Sanders will lose a state that he targeted to win on a night when Clinton is running up the delegate score in the South.

10:27 PM
Trump Alters The Paradigm

Behind an oh-so-presidential podium that read “Text Trump to 888022,” Trump just held a press conference, a move that’s a bit of a departure from the typical victory-speech format. With Chris Christie standing next to him, looking frankly a little like a glassy-eyed political spouse at a different kind of press conference, Trump hit his typical stump speech points — jobs going to China, the wall and how it will be paid for, and the threat of Syrian immigrants.

Why the press conference? One might call it a power move — part of Trump’s charm to his supporters is how unpredictable he is. He’s most comfortable when he’s speaking extemporaneously, and he’s also been thumbing his nose of late, in the wake of all his victories, at the media and all the nay-saying about the likelihood of a Trump presidency.

Many of the questions proffered by reporters were about whether or not Trump will actually stick to his on-the-trail policy offerings, to which Trump offered, yes, but also “I’m a common-sense conservative,” which came in a portion of the event when he was asked about his view on defunding Planned Parenthood (he wants to do it, but also offered that the organization provides a lot of useful health services to women).

One of the most overarching themes, however, was that the establishment just couldn’t discount him, not with the “millions and millions of people,” as Trump put it, who support his candidacy. He talked about how he was re-energizing the Republican Party, stealing support from Democrats, and generally changing the paradigm.

He sure isn’t wrong.

10:23 PM

Minnesota is probably not going to be won by Trump. With a little over a third of precincts reporting, Rubio is ahead there 34.5 percent to Cruz’s 30.4 percent. Trump is way back at 21.8 percent, and he’s not leading in any congressional district.

10:19 PM

Let’s just remember that a contested convention could be a costly thing to the party. Costlier than nominating Trump? Well, that’s not for me to decide. But you’d probably rather have a 25 percent chance of beating Trump cleanly than a 35 percent chance of a contested convention, to throw some semi-arbitrary numbers out there.

10:19 PM

Nate, what do you make of this:

10:16 PM
Alabama Trump Alabama Clinton
Georgia Trump Arkansas Clinton
Massachusetts Trump Georgia Clinton
Tennessee Trump Tennessee Clinton
Virginia Trump Texas Clinton
Oklahoma Cruz Virginia Clinton
Texas Cruz American Samoa Clinton
Alaska Not called Oklahoma Sanders
Arkansas Not called Vermont Sanders
Minnesota Not called Colorado Not called
Vermont Not called Massachusetts Not called
Minnesota Not called
Projected winners so far

Source: ABC

10:14 PM

Trump said he’s doing well among Latino voters, citing Nevada exit polls. Why cite numbers from a week ago? Because the Republican electorate in almost all the states that have voted tonight is heavily white — so much so that in the nine states for which CNN has released exit polls, only in Texas is there enough data to say how Latino Republicans have voted (Cruz won a plurality of their vote) and in no states were there enough black voters for meaningful data.

Alabama 1% NA% NA% NA% 4% NA% NA% NA%
Arkansas 1 NA NA NA 3 NA NA NA
Georgia 2 NA NA NA 7 NA NA NA
Massachusetts 2 NA NA NA 2 NA NA NA
Oklahoma 2 NA NA NA 1 NA NA NA
Tennessee 1 NA NA NA 3 NA NA NA
Texas 10 26 32 26 3 NA NA NA
Vermont 1 NA NA NA 0 NA NA NA
Virginia 2 NA NA NA 9 NA NA NA

Source: CNN

10:13 PM

10:12 PM

As Dave said, Rubio is in danger of not hitting statewide delegate thresholds in Alabama, Tennessee and Texas. He needs to hit 20 percent in all three, but he’s only at 17.8 percent in Alabama, 19.7 percent in Tennessee and 18.4 percent in Texas. It would not be good for him if those numbers hold.

10:07 PM
100401ANNE_LIAnne Li

Chris Christie just called Trump’s campaign a “movement,” echoing Sanders’s frequent calls for a political “revolution.” Although dramatic calls for change are not unusual in elections, such rhetoric made me think about why candidates in both parties are calling for such sweeping change. The economy is actually doing pretty well and has improved since the Great Recession, but Americans still perceive it to be bad.

Other than the economy, maybe there are different reasons for the unrest in each party. Democrats see injustices suffered by people of color, women and the middle class, while Trump supporters are watching the Islamic State group, undocumented immigration and the legalization of gay marriage and feel like their values are under attack. It’s one of the main reasons why, though there are some similarities between the Trump and Sanders campaigns, you really shouldn’t lump them together.

10:04 PM

10:04 PM

Maybe Rubio will win a state tonight after all. With about 10 percent of the vote counted from the Minnesota caucuses, he has 36 percent of the vote, compared with 29 percent for Cruz and 21 percent for Trump.

9:59 PM

Yeah, Trump has certainly siphoned off some traditionally Democratic voters, but he’s leading a cult of personality movement — I’m not totally sure they would vote Republican down the ballot.

9:59 PM

Well, turnout is up. It’s too early to say if he’s expanded the party. Party-building is real work — it involves fundraising for other candidates, registering voters. Nothing he’s done so far has the mark of something lasting.

9:58 PM

So in his remarks tonight, Trump said he has expanded the Republican Party? Is there evidence for that?

9:58 PM

Amid all the arguments about which candidate “beat expectations,” it’s worth comparing each candidate’s vote share tonight to their pre-election polling averages, as calculated by FiveThirtyEight. On average, in the eight states where we published polling averages ahead of time — Alabama, Arkansas, Georgia, Massachusetts, Oklahoma, Tennessee, Texas, Virginia —  Cruz is beating his polling numbers by 4.4 percentage points. Rubio is running just slightly ahead of his polls as well. Trump’s numbers are right in line with where polls had them, meanwhile.

Trump 37.7 38.4 +0.7
Cruz 20.1 24.5 +4.4
Rubio 19.9 22.3 +2.4
Kasich 7.0 6.9 -0.1
Carson 7.4 6.0 -1.4
Cruz beating his polls

Based on FiveThirtyEight polling averages and results reported as of 9:45 p.m.

This is consistent with evidence from the exit polls showing that Trump didn’t close especially well; most late-deciders went to Cruz and Rubio instead. Still, Trump started from far enough ahead to hold on for the win in the vast majority of states.

9:57 PM

From our friends at ABC News:


  • Alabama – Trump
  • Georgia – Trump
  • Massachusetts – Trump
  • Tennessee – Trump
  • Virginia – Trump
  • Oklahoma – Cruz
  • Texas – Cruz
  • Uncalled: Alaska, Arkansas, Minnesota, Vermont


  • Alabama – Clinton
  • Arkansas – Clinton
  • Georgia – Clinton
  • Tennessee – Clinton
  • Texas – Clinton
  • Virginia – Clinton
  • American Samoa – Clinton
  • Vermont – Sanders
  • Oklahoma – Sanders
  • Uncalled: Massachusetts, Colorado, Minnesota
9:54 PM

Trump said in his news conference that he beats Clinton in several head-to-head polls and suggested that in a matchup against her, he’d have a better chance than Cruz, who in turn would do better than Rubio. The reality is almost exactly the opposite: Rubio is essentially tied with Clinton — in hypothetical, premature, not particularly worthwhile polls — while Cruz on average trails Clinton by 4 points and Trump trails her by 5. Then again, this is hardly the first time Trump has cited only the polls most favorable to him.

9:48 PM
Rubio Is In Danger Of Missing Key Delegate Thresholds In Several States

If there’s a theme for Rubio so far tonight, it could be that he’s falling tantalizingly short in multiple places and probably wishes he could redistribute some of his voters across state lines. It looks as if he’ll just barely fall short of winning Virginia, where he’s on track to lose to Trump by just 2 points. But the bigger suspense is whether he will meet thresholds to win statewide delegates in places like Georgia, Tennessee, Texas and Vermont.

Interestingly, Rubio seems to be well above 20 percent in Arkansas and Oklahoma, and he doesn’t really need to be: The thresholds there are just 15 percent. However, the really bad news for Rubio backers: He’s winning 19.5 percent of the vote so far in Vermont, 18.9 percent in Texas, 16 percent in Alabama and 19.9 percent in Tennessee. In all those states, the thresholds are 20 percent. His best-looking state with a 20 percent threshold right now is Georgia, where he’s taking 22 percent.

Things could change, but Rubio better hope he can up his game in late returns from these crucial states. Otherwise, he could trail Cruz substantially in delegates at the end of the night, despite winning a comparable share of the national vote.

9:43 PM
100401ANNE_LIAnne Li

We were lying when we said we hoped you would stay with our live blog all night — we want you to take a quick break and sign up for our 2016 newsletter. You should also subscribe to our political podcast: FiveThirtyEight Elections. So, stay in touch! With so many options, you have no excuse not to.

9:41 PM
Starting To Think About The New Rules

It appears that four candidates (even Kasich!) might be able to walk away from tonight telling a story to themselves about why they are in good shape. I’m not even going to try to parse who’s right, but something Nate said in the last podcast has been rattling around in my head all evening.

More and more, I think we’re going to have to have a “new rules” reckoning. And if Trump wins the nomination, we can’t go into the general election pretending that more conventional elections offer any reliable lessons. Here’s Nate:

Every 40 years, you get a new set of rules. But maybe Trump’s even rarer than that. Maybe it’s kind of a “once every 200 year” phenomenon.

The Donald Trump thing is so momentous and so unusual. All bets are off, but … you almost have to say “the sky is orange” and things are changing, and they’re weird, and this is extremely consequential. You can’t go back [in the general] and say, “We’ll go back and pretend that never happened, and now we’ll just pretend that Donald Trump is Mitt Romney all over again and all the old rules apply.” Some of them will, some of them won’t, but you at least have to think about that.

If you want, you can listen to our Super Tuesday preview podcast below. Much of it is still relevant! But we’ll also have a new episode reacting to tonight’s results for you tomorrow.


Subscribe: iTunes | ESPN App | Download | RSS | New to podcasts?

9:36 PM

One of the states that Sanders had hoped to win is Massachusetts, but Clinton is leading by 5.2 percentage points there, with 35 percent of precincts reporting. That lead may shrink as more votes come in from the center and western parts of the state. Still, Clinton is slightly better than a coin flip to carry the state at this point.

9:36 PM
If This Were Any Other Year …

I hate to jump on the “in any other year” train but, well, I’m going to. I’ve heard at least one commentator say tonight, “if it were anyone else, we’d be referring to Trump as the presumptive nominee.” I think there’s a good chance that if the cast of characters were different, we’d probably expect Rubio to drop out tomorrow, especially as Cruz won in Oklahoma. Rubio just hasn’t won any contests (though obviously, we don’t know all of tonight’s results yet). There’s lots that could be said about this, but I think it speaks to two key features of this year’s nomination contest (other than, like, the fact that Trump is in it). First, the field has been stable and the winnowing has been slow.

The other thing is that the symbolism has quickly changed. Before the primaries, people talked about the “establishment lane.” While Trump has racked up victories and candidates, people seem to have stayed focused on who was in second. Or “strong third-place” finishes. The gap between the language used to describe the situation and the emerging mathematical realities suggests just how much informal processes – from the elite invisible primary to the expectations game in early contests – typically overshadow the formal ones. And the thing about informal rules is that they can sometimes be pretty easy to break.

9:31 PM

Betting markets think it’s been a very good night for Cruz, who was just called Oklahoma’s winner by CNN, to go with his victory in his home state of Texas. His chance of winning the Republican nomination is up to 5 percent according to Predictwise, higher than it’s been since early February. Trump has slipped slightly to 81 percent, while Rubio is at 11 percent. Clinton’s chance of winning the Democratic nomination has held steady tonight at 96 percent, and she’s at 64 percent to win the presidency.

9:30 PM

Several people on Twitter note that although Texas and Oklahoma voted together on the GOP side, they split in the Democratic contest, with Texas going to Clinton and Oklahoma to Sanders. So maybe they aren’t so identical after all. (By the way, to those saying less-than-printable things: I used to live in Texas and met my wife there. My line about them being the same place was a joke!)

9:29 PM

The Cruz win in Oklahoma is surprising not just because it pumps the brakes a little bit on a big Trump night, but also because it turns on its head some conventional wisdom that has been circulating about Cruz’s campaign for the past couple of weeks — that its dirty-tricks reputation (rest in peace, Dick Nixon) was having an adverse effect on its evangelical support, and that those voters were going over to Trump and Rubio. There were rumors about rumors that the Cruz team had supposedly spread in Iowa about Carson dropping out, and then there was a false video about Rubio denigrating the Bible that a Cruz staffer circulated, etc etc.

But according to ABC, three-fourths of Oklahoma’s Republican primary electorate were evangelical and nearly four in 10 of them went for Cruz.


9:27 PM

One thing I’d love to know about Oklahoma, which the networks have called for Sanders: how well he did with Native American voters. Unfortunately, there’s no breakout of this in the exit polls. Native Americans are lumped into the “other” category, and the sample size is too small for the exit poll to have reported the Clinton-Sanders split among that group.

9:24 PM

We still have a ways to go. Alaska, Arkansas, Minnesota and Vermont have yet to be called on the Republican side. On the Democratic side, it’s Colorado, Massachusetts and Minnesota. Obviously, there are still delegate totals to be determined on both sides.

9:22 PM
Clinton’s Enormous Lead Among Black Voters
Today’s Democratic contests have looked like previous ones in the Democratic race: Most voters age 18 to 29 have backed Sanders, and most African-American voters have cast a ballot for Clinton. Two reasons these two diverging (and overlapping) demographic groups generally add up to an edge for Clinton: (1) In many states there are more black Democratic voters than young ones, and (2) she’s winning among black voters by bigger margins than Sanders’s margin among young voters. In each of the seven of the nine states for which CNN has posted exit polls so far that have enough black voters to produce statistically significant numbers, Clinton won at least 75 percent of their votes. In six states, she won at least 80 percent. Remarkably, Clinton is winning nearly as high a percentage of black Democratic voters as President Obama won among all black general-election voters.
Alabama 47% 87% 10% 15% 36% 55%
Arkansas 27 88 12 18 38 62
Georgia 50 83 16 14 42 55
Massachusetts 4 NA NA 19 30 70
Oklahoma 14 75 22 12 19 80
Tennessee 31 82 12 16 32 66
Texas 16 80 18 21 32 66
Vermont 1 NA NA 15 5 95
Virginia 26 84 16 16 30 68

17-year-olds who will be 18 by the general election can vote in Vermont and Virginia and are included in 18-29 vote numbers

Source: CNN

9:20 PM

As the night progresses, it’s looking better and better for Ted Cruz and a little less rosy for Trump and Rubio — relative to expectations. Thanks to the Sooner vote, Cruz has extra justification to stay in the race a lot longer. Trump isn’t getting the sweep some thought possible and will lose at least two states. Rubio is doing better than he has in previous primaries but is teetering on the edge of important delegate thresholds in Texas, Tennessee and Georgia. Also, Rubio’s hopes of finishing ahead of Cruz look bleak in many Southern states.

9:19 PM

I mentioned earlier that Rubio’s support is hard to pin down. He tends to win those areas that don’t fit into the obvious strengths of the other candidates, rather than necessarily having areas that are all that strong for him on his own. That leads to a lot of second-place finishes, as he had in Virginia tonight.

The upside case for Rubio is that he also has the broadest appeal among Republican voters; he generally has the highest favorability ratings of any Republican candidate, for example. And in some polls — not all, but some — he runs even with or slightly ahead of Donald Trump in a hypothetical one-on-one matchup.

The problem is that Rubio may never be able to narrow down the field unless he can compel other candidates to drop out. And it’s hard to do that without having simple arguments he can make that can cut through the media filter. Winning states could do that, while close seconds probably won’t, especially given that the media has become increasingly cynical about Rubio’s expectations-spinning.

9:17 PM

Cruz’s (apparent) win in Oklahoma is a bit of a surprise based on the polling, but from an economic perspective, it makes sense that Texas and Oklahoma would vote the same way. Both are big oil states, with energy and mining making up a bit under 15 percent of their economic output. In part because of that, both came through the recession in pretty good shape; Oklahoma’s unemployment rate topped out at 7.1 percent, well below the nationwide peak of 10 percent. But with oil prices now slumping, both are starting to see their economies slow down. They’re practically the same state! (Just don’t tell them I said that.)

9:16 PM

NBC News has called Oklahoma for Sanders. That’s his second win of the evening, and it’s not a huge surprise. The state has very few black voters in it, but a win is a win.

9:14 PM
Clinton’s Speech: Groups And Regions

There are serious contrasts between Clinton’s speech and the one that Sanders gave for supporters in Vermont. While his was thematic, hers focused on some key regions — corporate misbehavior in the competitive state of Wisconsin (where I am writing this), voter disenfranchisement in North Carolina (where Obama won in 2008 but not 2012), and the suffering of the community in Flint, Michigan. I wouldn’t say her address was short on motivating ideas, but it was very firmly rooted in the conception of the Democratic Party as a coalition of groups and regions.

9:13 PM

9:10 PM

ABC News and the Associated Press have called Oklahoma for Cruz. Not a big surprise given that his lead is 3.5 percentage points and growing. Needless to say, the polls that had Trump winning were wrong. Cruz has now won two states tonight.

9:03 PM

Sometimes when exit polls and pre-election polls disagree, it’s the pre-election polls that had it right. Pre-election polls in Massachusetts had Clinton pulling ahead in the last week of the campaign, while exit polls earlier tonight had Sanders up. Clinton leads by 5 percentage points with 11 percent of the vote in so far, however. Obviously, there are still a lot of votes out, although The Upshot’s models have Clinton narrowly favored based on the vote reported so far.

9:03 PM
Clinton’s Victory Speech (In Most Places)

A little color commentary, in the midst of all tonight’s numbers.

Clinton is giving her Super Tuesday victory (in most places) speech right now, and it’s pretty apparent that she is already looking down the road to the general election, with Trump in her sights as her opponent. The last couple of days have seen Clinton shift a bit in her language, unmistakably calling out Trump without actually using his name.

“We have to make America whole,” Clinton just said to her adoring crowd. “Fill in what’s been hollowed out.” She almost seemed like she stole some pages out of Kasich’s touchy-feely handbook: “What we need in America today is more love and kindness.”

And of course, this: “Instead of building walls, we’re going to break down barriers and build ladders of opportunity and empowerment.”

Welcome to the stump speech of the next eight months!

9:00 PM

Cruz is officially on the board. He has won his home state of Texas, according to ABC News. Clinton, not surprisingly, has won the state’s Democratic contest. The things to watch now are Cruz’s margin over Trump and whether Rubio can finish above the 20 percent threshold for statewide delegates.

8:53 PM

Early on, Rubio’s chances of hitting Texas’s 20 percent threshold looked favorable. So far, it looks like a decent night for Rubio, a good night for Cruz in the sense that he’s winning Texas, and an about-as-expected night for Trump. But, Minnesota still awaits — it would be a big disappointment for Rubio if he doesn’t win there, given his encouraging near-miss in Virginia.

8:52 PM

CBS News, NBC News and Fox News have called Virginia for Trump. That means, if the pre-election election polls are correct, Rubio’s one really good shot to win a caucus or a primary is Minnesota. In terms of delegates, it really doesn’t mean anything, though, because Virginia awards its delegates proportionally.

8:49 PM
At Least Cruz Has Texas

Texas holds a very special place in the Facebook Primary: It’s the only state where Trump and Carson aren’t completely destroying their Republican rivals in Facebook likes:


Texas, not surprisingly, is Cruz’s best state, and the senator leads in counties in and around Houston, his hometown and residence. Trump secures the most likes along the (extremely Hispanic!) southern border, while Carson claims the lead in most of the rest of the state.

The Democratic side is less evenly split:


Sanders has twice as many likes as Clinton statewide — actually quite a bit worse than his 3-to-1 national margin — but loses out in a few border counties. See likes for more states.

8:48 PM
Dispatch From The Great American Highway

Helllooooo, Super Tuesday! I’m late to all this because I got caught in a bad rainstorm while driving from Columbus to Cleveland, but I’ve been keeping tabs on the results coming in via radio (ain’t I old-timey?) and got a chance to hear what the various campaign surrogates have to say about tonight’s likely outcomes. John Podesta, Clinton’s campaign chair, was on NPR, and it was pretty interesting to hear him try to spin all the questions that kept coming his way about how Clinton is going to court Sanders’s youth vote, now that she’s the likely nominee. He sort of tried to play it as “well, Bernie built a nice little movement for himself,” but didn’t have much to say beyond that — it’s an enthusiasm gap on the Democratic side that will be something to watch for.

On the Trump side of surrogacy, Rep. Chris Collins of upstate New York was also on NPR, talking about his guy — and funnily enough, Collins’s guy was actually Jeb Bush up until about a week ago. (Oh my gosh, has it only been a week since all that happened???) I can’t imagine a more strange endorsement decision, but Collins said that both were executive types, and he employed some fairly Donald-esque language, talking about how President Trump would assemble one of the best Cabinets the U.S. has ever seen.

Radio remains one of our greatest mediums, is my take-away.

8:47 PM

Micah, I think most people thought Cruz would end up with more delegates than Rubio — Cruz will get a big bump from Texas. I think the bigger question is whether Cruz can win Oklahoma or even Arkansas. That would be a big deal for him.

8:46 PM

Is it possible that Cruz comes out of tonight with more delegates than Rubio? Was that always going to happen?

8:44 PM

We’re about 15 minutes from the polls closing in Texas. Here’s how Texas Republicans voted in 2008 and 2012:

2008 Texas Republican primary
2012 Texas Republican primary
8:42 PM

Fox News has called Virginia for Trump, while other observers I’m following on Twitter say it isn’t quite a sure thing yet. Rubio has made the result closer there than pre-election polls projected but trails Trump by about 5 percentage points with most of the vote counted. Still, most of the remaining vote is in Rubio-friendly Northern Virginia. It’s worth remembering, however, that Virginia’s delegate allocation is highly proportional, so winning the state matters a lot more for the media narrative than for the delegate math.

8:40 PM

Trump, by the way, is also performing well in those poorer Southern states. That’s less of a surprise given Trump’s strength among less-educated, blue-collar workers. Perhaps more surprising is his success in Massachusetts, which has the most educated population in the U.S. by many measures. But western and central Massachusetts look very different from the wealthy, educated areas around Boston. There are plenty of people there who might be more receptive to Trump’s message.

8:37 PM

8:35 PM
Trump Comes In Fourth (!!) Place At Liberty University

In a surprising rejection of Liberty University President Jerry Falwell Jr.’s endorsement, Trump has apparently finished in fourth place in Liberty’s precinct in Lynchburg, Virginia. Lynchburg’s precinct 302 voted for Romney over Obama by a stupendous 3,120 to 153 in 2012. Yet tonight, Rubio appears to have won it with 513 votes to 387 for Cruz (who launched his campaign there), 162 for Carson, and just 90 for Donald Trump.

It appears that the profane, casino-owning Trump just didn’t play so well in one one of the most famously evangelical Republican precincts in the country.

8:34 PM

As expected, Clinton seems to be winning the Southern states comfortably, at least based on the results we’ve seen so far. Much of the attention has focused on Clinton’s strong support among African-American voters, and with good reason — at least in South Carolina, black voters provided her with much of her margin of victory. But it’s worth noting that Southern states such as Alabama, Georgia and Tennessee are also among the poorest states voting tonight and have experienced some of the slowest economic recoveries. (South Carolina, which Clinton also won in a landslide, was a similar story.) Given Sanders’s focus on economic issues, and his apparent support among low-income voters, it’s interesting to see Clinton do so well in these states.

8:33 PM

Don’t be fooled by all the Trump purple on this map: Rubio is performing impressively in Virginia’s urban crescent, stretching from Hampton Roads through the Richmond suburbs to Northern Virginia. Trump is currently ahead 36 percent to 31 percent statewide and is likely to hang on, but his margin is likely to narrow significantly once more of Fairfax County (in the D.C. suburbs) reports votes.

2016 Virginia Republican primary results
8:33 PM

How could Trump win so easily in states seemingly as dissimilar as Alabama and Massachusetts? Well, it’s complicated. In contrast, to the Democratic race, which falls along fairly predictable demographic lines, the coalitions within the Republican Party this year are unusual. But in running various correlations when preparing our state-by-state delegate targets, we found the following:

  • Trump does better in areas with lower socioeconomic status, and especially lower education status.
  • Trump does better in areas with a lower evangelical population.
  • Trump does better in areas with a high degree of racial animus, as measured by racist Google searches.

Put another way, Trump has several different groups of voters he does comparatively well and poorly with. Cruz’s coalition of evangelical and highly conservative voters is simpler by comparison, whereas Rubio tends to pick up the groups that the other candidates leave over.

Also keep in mind that the demographics of Republicans in a state may differ from those in a state as a whole. Our research — based on exit polls — suggests that many Republican voters in Massachusetts tend to live in fairly “downscale” or post-industrial areas, even though the state is fairly well-off overall.

8:31 PM

So far, Trump is winning where we expected him to win — ABC News has called Georgia, Alabama, Massachusetts and Tennessee for him — and we expected him to have a good night. But it doesn’t seem to me to be any sort of blowout that suggest the race ends tonight. We’ll see if that holds as the results keep coming in. The GOP race is closer than I think most expected in Vermont or Virginia, and it seems Cruz has a legitimate shot to win in Oklahoma.

On the Democratic side, I think the real story is Clinton romping with black voters. ABC News has called Clinton wins in Arkansas, Alabama, Virginia, Georgia and Tennessee, and a Sanders win only in Vermont, so far. If Sanders can win in Massachusetts and Vermont, it would be a moral victory. But in reality, Clinton will have a nearly insurmountable delegate lead by the end of the evening.

8:31 PM

Harry, can you give us a 30,000-foot overview of the Democratic and Republican results so far? Surprising/not surprising? Enchanting/not enchanting?

8:26 PM
Trump's Strongholds

Trump’s projected wins in Alabama, Georgia, Massachusetts and Tennessee come in three of the states where he could expect to have the best chance in the general election if he were to win the Republican nomination. We’ve written several times that Trump isn’t terribly popular nationally, based on how many voters approve or disapprove of him. According to Morning Consult online polls of more than 25,000 registered voters nationally — Democrats and Republicans — so far this year, in the median state, the share of voters who disapprove of Trump is 12 percentage points higher than the share who approve of him. But he does better than that in Alabama, Tennessee and Georgia — particularly in Alabama, one of just eight states nationally where Trump has a net-positive approval rating. He’s also relatively popular in three other states voting tonight: Alaska, Arkansas and Oklahoma. (Cruz, who represents Texas, is looking competitive in neighboring Oklahoma.)

Many of the states that look best for Trump are in the South, including Kentucky and Louisiana, two of the states other than Alabama where Trump has more approval than disapproval. “It is kind of amazing when you think about it,” CNN’s Dana Bash said, that a “brash billionaire from New York” is doing so well in the South. And it must be discouraging to Cruz, who spent so much time earlier this year in the southern states voting tonight.

Morning Consult also asked voters about Clinton, and her results aren’t terribly encouraging for the general election. Her median net approval in a state is -16 percentage points, four points worse than Trump’s. The good news for her is that some of her best states have some of the most electoral votes: New York, California, Florida, Illinois, Ohio and Pennsylvania. Trump is below his average in four of those states and below Clinton’s in all six. A candidate who wins those six states is 63 percent of the way to the needed 270 electoral votes.

8:20 PM
Colorado Ads — Clinton v. Sanders

The Political TV Ad Archive has been tracking ad buys in select states. Sanders and Clinton have both been blanketing Colorado, but their messaging has been different. A key Clinton ad focuses on immigration and a girl worried about deportation, while one of Sanders’s heavy rotation ads addresses fracking and the environment. Both are likely to hit home with different (and overlapping) demographics. A 2014 report by Latino Decisions found that 63 percent of Latinos in Colorado knew someone who was undocumented, and 35 percent knew someone who had been detained or deported. Colorado environmentalists successfully blocked the expansion of a coal mine last year and are proposing 11 ballot measures to block fracking. Colorado has been a place for heavy ad buys by Clinton and Sanders: As of mid-February, Sanders had spent $1.6 million on ad buys and Clinton just shy of $750,000. The state has 66 delegates to offer the warring Democrats.

8:15 PM

Tennessee has already been called for Clinton, but both pre-election polls and exit polls show a very close race in Oklahoma between Clinton and Sanders.

Maybe this doesn’t seem surprising, but to our demographic model — which accounts for the liberal-conservative orientation of each state, its racial demographics, and the share of voters who live in rural areas — Oklahoma and Tennessee looked pretty similar, and both were thought to narrowly favor Sanders in a 50-50 race. Oklahoma is an idiosyncratic state, however, with a large Native American population and characteristics of both the Midwest and the South.

8:14 PM

It looks like Virginia Beach might save Trump from a loss to Rubio in Virginia. Trump has performed better there than he has in other Virginia suburbs — he’s currently ahead there 38 percent to 30 percent. Virginia Beach is home to a large military population, but also a high evangelical population.

2016 Virginia Republican primary results

8:10 PM

As we bite our nails tonight (and perhaps for the next couple weeks, as Nate suggested), the bigger question still looms: Why have this year’s nomination contests been so intense?

There are lots of possible explanations, but today I stumbled onto an intriguing one: because partisanship is so highly predictive of general election votes, the primaries are the big chance for intense competition over ideas and candidates to occur – nationally, per my Massachusetts and Oklahoma post. The general election will be a scramble for a turnout and for a few persuadable voters in a few close states. Perhaps the nomination race has become so much like the general used to be – high-profile, contested, prolonged – because the general has become more narrow and predictable.

8:08 PM

Just a reminder as more results roll in throughout the night: Earlier this week, we launched an interactive graphic showing how many delegates each candidate needs to win in every primary and caucus to take the nomination. The targets are based on more than just simple math; they take into account where candidates are most likely to do well and mark the most realistic path to victory for each one. Here was the state of the Republican race going into the day:

8:05 PM

Exit polls suggest the potential for a couple of upsets in the states that just started reporting results. Oklahoma exit polls have Cruz narrowly ahead, with Rubio and Trump tied for second, when pre-election polls had shown Trump with a reasonably clear lead.

Meanwhile, in Massachusetts, exit polls have Sanders narrowly ahead of Clinton, when pre-election polls had shown Clinton surging and pulling ahead of Sanders.

8:05 PM

8:00 PM

ABC News has called Alabama, Massachusetts and Tennessee for Trump. Again, none of those are big surprises. The key is to watch the 20 percent thresholds in Alabama and Tennessee for delegates that I spoke about in my Super Guide to Super Tuesday.

8:00 PM
Nate And Harry Discuss The Super Tuesday Stakes
7:54 PM

Surprise, surprise (OK, not really): Rubio has a Beltway fan club. In Virginia’s 8th congressional district, situated mostly inside the Washington Beltway, Rubio is currently crushing Trump 46 percent to 19 percent, by far his best showing anywhere in any primary in the country so far. The even better sign for Team Rubio: There’s a ton more votes in that part of Virginia to be reported, meaning that he’s likely to narrow his current 37 percent to 32 percent statewide deficit.

7:51 PM

Why is Trump winning Georgia easily — our ABC News colleagues just called the state for him — while Virginia is extremely competitive between Trump and Rubio? One reason is the higher share of college-educated voters in the primaries in Virginia: About 60 percent compared with 50 percent in Georgia, according to exit polls. Still, this doesn’t explain the whole difference: Trump is winning college-educated voters in Georgia but losing them in Virginia. That some Republicans who are active in party politics live in Virginia — maybe you can think of them as literal members of the establishment — may also matter at the margin.

7:47 PM
Red-Meat Bernie Bros

Oklahoma may be the reddest state in the nation, but its college students are still crazy about Sanders. Excluding Carson, Trump has the most Facebook likes statewide among the remaining contenders (of both parties), but Sanders tops him in Cleveland and Payne counties, the respective homes of the University of Oklahoma and Oklahoma State.


But it’s not as though these counties are pillars of liberalism: In the 2012 general election, Romney won Payne County by 28 percentage points and Cleveland County by 26 points.

7:46 PM

Georgia looks like another South Carolina on the Democratic side. Clinton is winning early with nearly 70 percent of the vote, and very little of metro Atlanta or the black belt in the southwestern part of the state has reported. It wouldn’t be surprising if Clinton got over 70 percent of the vote in the state.

7:45 PM

Clinton has won the American Samoa Democratic caucus, according to ABC News.

7:44 PM

Sanders condemned super PACs in his Vermont victory speech tonight, as he does on most public occasions, so it’s worth noting that there is a super PAC that’s backing him, just as there is for all the other major candidates. The group, National Nurses United for Patient Protection, has raised $2.3 million, according to its most recent federal filing.

7:43 PM

We’ll also start getting results in Massachusetts soon; here’s what happened in the state’s last competitive Democratic primary:

2008 Massachusetts Democratic primary
7:42 PM

And here are the 2008 Democratic results for Oklahoma:

2008 Oklahoma Democratic primary
7:41 PM

Be cautious of Trump’s 39 percent to 29 percent lead over Rubio in Virginia so far. Most of that is from the western part of the state; there are not a lot of votes in from Rubio-friendly Northern Virginia yet.

7:39 PM

7:39 PM

Sanders thanked Vermont voters — those who know him best, he said — for giving him at least one win tonight. He really is popular at home: He has the highest approval rating among home-state voters of any U.S. Senator, according to Morning Consult polls conducted last year. Some 83 percent of Vermonters approved of Sanders, 5 percentage points more than runner-up Susan Collins of Maine. Ted Cruz was down at 52 percent in Texas, which may help explain why he is expected to have a much tougher time holding his home state tonight. Rubio had the approval of just half of voters in Florida, which votes in two weeks.

7:37 PM

Oklahoma is a useful state if you want to understand this year’s races. Polls close in Oklahoma at the top of the hour, and here’s how the state voted in the Republican primaries in 2008 and 2012:

2008 Oklahoma Republican primary
2012 Oklahoma Republican primary
7:35 PM

Exit polls had Clinton beating Sanders by 25 percentage points in Virginia and 31 percentage points in Georgia, which is pretty close to pre-election polls in both states. Clinton’s margin among actual votes so far is larger in both states, but there aren’t many votes counted yet.

7:33 PM
Race And The Race

This week, a political race that has long flirted with the description “cray cray” just got a bit crazier, when Trump failed to disavow white supremacist David Duke and the KKK on CNN. In the ensuing firestorm of reframing and backpedaling, GOP heavy-hitters including Mitch McConnell and Paul Ryan condemned Trump.

Many people have called Trump racist; many others, from former Arizona Gov. Jan Brewer to Heisman Trophy winner and former NFL player Herschel Walker, stepped up to defend the billionaire. First of all, we might do well to shift the debate away from labeling Trump to analyzing his behavior in a political science context. (More broadly, that’s what commentator Jay Smooth describes as “the difference between the ‘what they did’ conversation and the ‘what they are’ conversation.”)

From a political science perspective, racial appeals in campaigns work to activate voters, particularly certain segments of voters. Research by Princeton professor Tali Mendelberg and, later, by Yale professors Gregory Huber and John Lapinski finds that racial appeals motivate voters. Mendelberg thinks implicit appeals — sometimes called “dog-whistles” — work when voters reject blatant racial appeals as unseemly. The Yale professors ran an experiment with 6,300 participants arguing that “implicit appeals are no more effective than explicit ones in priming racial resentment in opinion formation.” (More highly educated Americans are less likely to respond to these appeals.)

Another body of research shows that “racial resentment” rises sharply for white voters who are in economically distressed areas — ones with high unemployment rates and, particularly during the Great Recession, home foreclosure rates — and ones who live in close proximity to black Americans. (Interesting, the same was not true for proximity to Latino Americans.)

All of which is to say: A politician like Trump who has not been shy about making racialized statements does have a stake in invoking dog-whistle politics and appealing to white Americans who feel economically threatened and respond to racial appeals.

7:28 PM
Republicans In Massachusetts, Democrats In Oklahoma

People are obviously getting restless about the relationship between primaries and the general election. Are the primary results predictive? Does high turnout in the primaries and caucuses mean high turnout in the general?

I’ve been especially preoccupied with the apparent importance of New Hampshire and Iowa for Republicans, given that both states seem to me pretty reliably in the blue column for presidential elections. I’m also following the Democratic race today in Oklahoma, a state where Mitt Romney won nearly 67 percent of the vote in 2012.

Nomination influence by states that don’t deliver for the party in the general has a long and important history. In the decades after the Reconstruction, Southern states played a non-trivial role in nominating Republican candidates, even though they never delivered any Electoral College votes for the South.

As we think about polarization and primaries, this illustrates how informal, symbolic ways of understanding primaries can sometimes work the same way as delegate math. Just as Republican hopefuls needed Southern delegates in the 1890s, tonight Clinton is counting on a strong showing in Southern primaries not just to garner delegates, but also to meet expectations about her strength with minority voters. We’ve started to see this a bit with the Democratic candidates’ efforts to showcase the concerns of African-American voters – with Sanders showing some strength in white, rural areas, she needs those primary votes to win, even if the Democrats stand little chance to do well in those states in November.

7:22 PM

Here’s a forecast for what will happen if four different candidates win states tonight, as very early exit polls suggest might be possible, while Trump still emerges with the clear majority of states and delegates: None of us are going to get much sleep between now and March 15. Trump will look more stoppable than when the night began, but the plan for stopping him will be less clear than ever and may increasingly involve talk of a contested convention.

7:21 PM

The first real votes out of Virginia are from Chesterfield County, and they show Trump at 37 percent, Rubio at 33 percent and Cruz at 19 percent. That could be a good sign for Rubio that he’s not going to get blown out. Chesterfield is a bit of a GOP bellwether: It’s the well-educated Richmond suburbs, but it’s not as wealthy or well-educated as Henrico County to the north. We’ll see.

7:19 PM

Right now, the updated exit polls have Kasich ahead in Vermont and Rubio dead-even with Trump in Virginia. Given that pre-election polls favored Cruz in Texas, it seems quite possible that four candidates will win on the Republican side tonight. If that happens, I don’t think we’ll be seeing a winnowing of the Republican field any time soon.

7:15 PM

The most interesting thing to me in the early exit polls (if you believe them, and you should be careful with them) is that Trump is really only ahead in Georgia. He’s at best tied in Vermont with Kasich and perhaps slightly ahead of Rubio in Virginia. He’ll do better in Alabama and Tennessee, based on pre-election polls, but I think the idea that Trump is running away with this thing isn’t self-evident. At least, not yet.

7:14 PM
Rubio Has Led Very Few Polls

Harry mentioned that Rubio has trailed in almost every state poll. He’s right, especially recently. I checked the 239 state Republican polls that HuffPost Pollster has compiled for 22 states. Rubio led in just five — and only two since June 2013. In 18 of the 22 states, he has never led in a poll. Three of the four states where he has led at least once vote today: Minnesota, Virginia and Texas. (The fourth is Florida, Rubio’s home state; HuffPost Pollster hasn’t compiled Utah polls.)

The picture isn’t much better nationally for Rubio: He has led in just 22 of 307 national polls that HuffPost Pollster has compiled, and none since a YouGov/Economist poll in June 2015. He has trailed in 206 national polls since then.

7:13 PM
‘Berners’ Out Front

As FiveThirtyEighter Matthew Conlen tweeted, some Bernie Sanders supporters have camped out in front of our office building this evening:

They must be huge FiveThirtyEight fans! Or… maybe they’re just trying to get some attention from WABC, the local ABC affiliate that shares our building (actually, vice versa). The channel’s live studio is visible from the sidewalk where the Sanders crew set up shop. In fact, a couple of them — Ting Barrow, 71, a retired webmaster and Ned Kaufman, 63, a contractor who works on neighborhood preservation projects — told me as much. Both of them feel like the media needs to cover Sanders more, and they’re trying to inspire such coverage. I guess I took the bait!

Ting Barrow (left) and Ned Kaufman (right) picket for Bernie Sanders outside FiveThrityEight's offices

Ting Barrow (left) and Ned Kaufman (right) picket for Bernie Sanders outside FiveThirtyEight’s offices

Ritchie King

I’ve actually been comparing the number of news articles written about each of the candidates over the course of the primary season, and it turns out, when it comes to sheer volume, the media has been giving Sanders a pretty fair shake.


But what about tone? Kaufman sees a tendency among news outlets to talk about Clinton as if she’s the only candidate. And he had a pretty good example on hand. On the cover of today’s New York Times, a story specifically about the Clinton campaign’s plan for beating Trump in the general election was headlined: “Democrats prepare a Trump Battle Plan.”

For comparison, that same story on The New York Times website is headlined “Inside the Clinton Team’s Plan to Defeat Donald Trump.” To be fair, it could have been a space-saving maneuver (real estate on A-1 is precious!), but it could also have been the Times accepting that Clinton looks increasingly likely to take the nomination and move on to the general (maybe a mite prematurely). The question of whether Clinton regains the status of “inevitable nominee” is one that could very well be answered tonight.

7:11 PM
Rubio Wins In Literal Backyards Of Republican Establishment

According to preliminary exit polls, here’s how the vote broke down in the D.C. suburbs of Virginia:

Rubio 40

Kasich 24

Trump 18

Cruz 13

7:08 PM

7:08 PM

From looking at the gender breakouts in the version of the exit polls, here’s roughly what they show in the Republican race.

Vermont: Trump 32, Kasich 31, Rubio 19
Virginia: Trump 34, Rubio 31, Cruz 16
Georgia: Trump 40, Rubio 23, Cruz 22

Closer than expected in Virginia and Vermont, although there wasn’t much polling in Vermont on the GOP side. The early exit polls are reasonably close to the pre-election polls in Georgia, however.

7:02 PM

NBC is saying Virginia is too close to call on the Republican side. Note: That’s different than “too early to call.”

7:00 PM

ABC News has called Georgia and Virginia for Clinton and Vermont for Sanders. No surprises there. The FiveThirtyEight polls-only forecasts favored Clinton by 36 percentage points in Georgia and 24 percentage points in Virginia. Meanwhile, the Vermont polls-only forecast gave Sanders a 77-percentage-point edge.

6:55 PM

Soon we’ll be counting actual votes, and we won’t have to worry about how candidates are “setting expectations” with the press. But as you read articles like this one and this one — Rubio’s campaign at once seems to be lowering and raising expectations — it’s worth remembering that candidates sometimes have contradictory goals in mind.

On the one hand, there’s a fair amount of research to suggest the media sometimes judges candidates for how they perform relative to expectations, and especially relative to pre-election polls. A candidate who gets 27 percent of the vote when polls had him with 20 might be deemed to have a successful evening, for example, and the “momentum” the more favorable press coverage produces may actually help him in future states. That would argue for lowering expectations whenever possible.

On the other hand, a candidate who lowers expectations too much may risk lowering the morale of his supporters or having them cast ballots for other candidates. That can be risky too, especially in multi-candidate races.

So sometimes candidates will get caught doing both expectation-raising and expectation-lowering at once. Also, sometimes campaigns can be every bit as surprised by the results as reporters and number-crunchers are in the event of a last-minute shift.

6:53 PM

Q: When is the last time Marco Rubio led in any state poll? — commenter Joshua David

A: Oof, that is a great question. He definitely led in Utah in February among Republicans.

I don’t have a list of polls from each state. But of the many state polls taken across the country, he’s trailed in almost every one. The two states where he has been competitive, according to the data I’ve seen, are Minnesota and Utah. That includes what I believe was his last lead in a poll: A Utah survey putting him barely ahead of Cruz.

6:52 PM
Trump’s Hold On Appalachia

With the exception of Carson, Trump has by far the most Facebook likes nationwide among the Republican candidates; in most states, he has the greatest share of likes in every single county. Virginia is a rare exception: Trump leads in 94 counties (and all 38 of Virginia’s independent cities), but Rubio leads in Arlington County!


Trump’s lead in Virginia overall isn’t as exceptional as his dominance in the western region of the state, where Trump outpaces his national share of likes by around 20 percentage points. A similar pattern can be seen in eastern Kentucky, southern West Virginia and northeastern Tennessee — the core of Appalachia is lining up behind Trump. See likes for more states.

6:50 PM
Didn't [Insert State] Used To Vote On Super Tuesday?

Probably. The list of participating states has fluctuated since the first Super Tuesday, in 1980. Forty-one of the 50 states have held their vote on a Super Tuesday at least once, and the Super Tuesday with the highest number of participating states and territories, 25, took place in 2008.

6:46 PM

We’s about 15 minutes from the polls closing in Virginia — here’s how Republicans voted there in the 2008 and 2012 primaries:

2008 Virginia Republican primary
2012 Virginia Republican primary
6:46 PM

Q: What is a good/great/bad night for Hillary in delegate count — what is her/Bernie’s floor/ceiling? — commenter Rodrigo Paramo

A: Our delegate tracker interactive projects that Clinton would need to win 453 delegates tonight to be “on track” for a majority of pledged delegates, while Sanders would need 412. So, I’d say anything under close to 500 would be a very good night for Clinton, while anything over 550 would be a terrific night. Anything over 400 would be a celebration-worthy night for Sanders, while anything under 300 would be a viability-devastating outcome for him.

6:42 PM
Clinton Closes The Facebook Gap In The South

We’ll have real results soon, so don’t put too much stock in this — but how do Clinton and Sanders compare on Facebook in the states voting today? Sanders has nearly three times as many likes as Clinton nationwide, but his margin is a lot narrower in today’s “SEC” primary states. Clinton enters Super Tuesday with a distinct demographic advantage.

Arkansas has just the 13th highest share of African-Americans in the U.S., with a smaller percentage than New York or Delaware. However, it was also Clinton’s home from 1974 to 1993, and she leads Sanders in Facebook likes in 13 counties. Her support is largely concentrated in the southern half of Arkansas, which is disproportionately black:


She also performs relatively well in Alabama, winning eight counties:


And she takes six in Georgia:


There’s a lot of purple on these maps, but those three states are still big outliers for Clinton — in 42 states, she wins zero counties against Sanders. Clinton led in seven South Carolina counties and went on to win the primary by nearly 50 percentage points. See likes for more states.

6:41 PM

Nate — Giving Trump a 34 percent chance to win if he’s the nominee seems pretty consistent with what betting markets indicate. Predictwise, which uses betting data to derive implied probabilities of election outcomes, shows Trump with an 86 percent chance of winning the nomination and a 26 percent chance of winning the general election. That suggests he has a 30 percent chance of winning in November if he’s the Republican nominee. (That also means bettors think there’s a one-in-four chance of the U.S. electing President Donald J. Trump.)

6:38 PM

Julia — The problem with everyone staying in, at least according to the really cool New York Times delegate calculator thingamabobber, is that states become increasingly winner-take-all or winner-take-most from March 15 onward, so a plurality of the vote for Trump could go a long way.

Of course, if these candidates are actually winning states themselves — Kasich in Ohio, Rubio in Florida — it might be a little different. The bottom line is that stopping Trump is going to require candidates to start winning states sooner or later. If and when they start doing that, you could debate whether everyone-stays-in or everyone-else-drops-out is the better approach.

6:32 PM

We recently discussed these kinds of exit poll questions in response to a question from one of our podcast listeners (the conversation starts at the 47-minute mark). As Nate explained, one thing to keep in mind is that voters tend to find a candidate they like and then offer answers to poll questions accordingly. So if you like Trump, you’ll find yourself saying that you want a political outsider. If you like Clinton, you’ll want political experience. So looking at the numbers above, it’s not that Alabamians love outsiders — it’s that they really like Trump.

6:30 PM
Trump’s General Election Odds: A Twitter-Sourced View

We’re getting way ahead of ourselves, I know, but I asked my Twitter followers earlier today what chance they thought Trump would have to win the general election, should he eventually win the Republican nomination. The crowdsourced answer is that Trump would have a 34 percent chance:

Seems reasonable enough, and pretty consistent with my back-of-the-envelope impressions. Trump would be an underdog, but there’s a lot we don’t know — and his becoming president would be well within the realm of the possible, especially if Trump’s opponents were complacent about the possibility.

It’s interesting how much disagreement there was among respondents on Twitter, however. About a quarter of my followers thought Trump had almost no chance. More than a third thought he’d be even-money or better in November, however.

Perhaps there’s not just uncertainty about what Trump’s chances would be, but also lots of philosophical disagreement about how we’d go about assessing them.

6:28 PM

According to FiveThirtyEight’s handy new delegate interactive, Trump needs about 225 delegates in tonight’s contests to be “on track” to win 1,237 — a majority — of convention delegates by June. By contrast, Cruz needs 302 delegates (mostly because his route runs through his home state of Texas), while Rubio needs only about 203.

However, given Trump’s dominance so far, winning only 225 delegates tonight would have to be considered something of a disappointment for his backers. What would a really terrific Trump night look like? I’d argue that he would need to win more than 300 of the 595 available to look dominant. What would an OK/unsurprising night look like? Probably between 250 and 300.

How about a disappointing night for Trump? That probably means anything less than 250, which would open the door for Rubio or someone else to overtake him in later winner-take-all states. Fellow data nerds: agree or disagree on these over/unders?

6:23 PM
Should Everyone Stay In To Block Trump?

Is this the best way to stop Trump — all four other candidates (Cruz, Rubio, Kasich and Carson) stay in the race?

A reader, Paul Winter, suggested this in our comments area, and so did former Bush administration official and Utah governor Mike Leavitt this morning on NPR.

The assumption is that this is the best way to deprive Trump of the majority of delegates to win the first ballot in the convention. This rests on the idea that the four candidates can win more votes together than any of them can individually. That flies in the face of all the conventional wisdom up until today — that the way to beat Trump would have been to coalesce around a single candidate.

The “stay in” logic suggests two things: that any candidate who drops out will shed voters to Trump as well as the other three opponents, and that beating Trump is no longer about a “show of strength” for another nominee, but rather about raw delegate counts, something we rarely see. And if the goal is to block a Trump majority, then the ultimate idea is to take the fight to the convention floor.

6:19 PM
Delegate Math

We write a lot about delegate math in this election — how it works against Cruz, for instance, or how it is front-loaded against Rubio. But browsing the delegate counts for the states that are voting today made me wonder something more basic: Why are the number of Republican and Democratic delegates different for the same state, and why isn’t the ratio the same? Oklahoma has more Republican delegates than Democratic ones, but Massachusetts has more than twice as many Democratic delegates as Republican delegates. What gives?

The short answer: Delegates are allocated by state based on a number of factors, including population but also the party lean of the state. The state’s party breakdown particularly matters in Democratic delegate allocation, and Massachusetts is far more Democratic-leaning than Oklahoma is.

The long answers for how states allocate delegates are really long: Thousands of words for Democrats and Republicans from The Green Papers.

6:14 PM

It’s a little early to reach sweeping judgments about patterns of where candidates are overperforming or underperforming their polls. But it’s at least worth noting that Bernie Sanders beat his polls by several points in New Hampshire, while doing quite a bit worse than his polls in South Carolina. That’s consistent with historical patterns, which show that candidates have tended to beat their polls in states where their state numbers exceeded their national numbers, as was true for Sanders in New Hampshire and Hillary Clinton in South Carolina. This is one reason that our “polls-plus” forecasts, which account for this tendency, can be slightly more aggressive than our “polls-only” forecasts.

6:12 PM
Tonight's Whole Foods/Cracker Barrel Indicator

In a few minutes, we will be served more election returns than we can possibly munch on in one night. But the data nerd (and inner foodie) in me will be eyeing an important indicator: the Whole Foods/Cracker Barrel divide.*

One of the reasons Trump is doing so well is that the Republican Party has lost so much ground in Whole Foods country, the well-educated, large metro suburbs where he is performing the weakest so far. In 2012, President Obama carried 75 percent of counties that now have a Whole Foods Market, compared with Bill Clinton’s 59 percent in the same counties in 1992. But more importantly, Trump is simply kicking ass in Cracker Barrel country.

In the four states that have voted so far, he has carried 81 percent of counties with a Cracker Barrel Old Country Store; they skew toward lower-income whites who lack college degrees. Republicans in Cracker Barrel country make up a far greater share of the GOP electorate than they used to: Mitt Romney carried 71 percent of counties with Cracker Barrels in 2012, up from George H.W. Bush’s 60 percent in the same counties in 1992.

Rubio has performed respectably in Whole Foods country thus far. He’s won three of eight Whole Foods counties so far: Polk County in Iowa and Charleston and Richland counties in South Carolina. If Rubio and Kasich were one super-candidate, they would have beaten Trump in two more Whole Foods counties as well: Greenville and Beaufort in South Carolina. But Trump’s dominance in Cracker Barrel land means Rubio badly needs to improve his shares tonight among the few “Whole Foods Republicans” who are left.

In other words, to exceed expectations tonight, Rubio needs to dominate the kale belt by winning places like Northern Virginia, suburban Atlanta, and the Twin Cities. And it wouldn’t hurt if Cruz stole a few plates of chicken and biscuits from Trump in places like rural Alabama and Tennessee.

*Fun fact: Conveniently, it just so happens that Super Tuesday covers both companies’ corporate headquarters: Whole Foods is based in Austin, Texas, while Cracker Barrel is based in Lebanon, Tennessee.

6:11 PM
Winner Take All?

If you’re reading this website, you’ve probably read about how Rubio’s plan is to stay competitive in the delegate count tonight, because most states are proportional, and then to win winner-take-all states like Florida and Ohio in two weeks to catch up to Trump. What you may not know is that some Republican Super Tuesday contests could become winner-take-all, or close to it.

As I detailed in my Super Guide to Super Tuesday, Republican delegate rules vary a ton by state. In a number of states, you can win all the district and statewide delegates by winning over 50 percent in each. Perhaps even more strangely, a candidate can win all the district/statewide in some states even without crossing 50 percent, so long as nobody else crosses that threshold. But will either of these scenarios happen?

Currently, no Republican candidate is near the 50 percent threshold in any state where that would result in winning all the delegates on the district or state level. Trump, however, looks like he could win almost all of the delegates in Alabama (50 delegates) and Vermont (16 delegates) because his opposition is divided. In Alabama, our polls-only projection has Trump at 43.4 percent while Rubio is at 20.4 percent and Cruz at 17.2 percent. The threshold to win any statewide delegates is 20 percent. In Vermont, the only poll taken in February had Trump at 33 percent, Rubio 15 percent, Kasich 14 percent and 13 percent undecided. The Vermont threshold to receive any delegates is also 20 percent.

Now, it’s likely Trump won’t back into a delegate sweep in these states, but clearly it is possible.

5:55 PM
Is Google Search Data Good News Or Bad News For Trump?

As Justin Wolfers notes for The New York Times, Google search traffic — an early sign of last-minute surges for Cruz and Rubio in Iowa and for Kasich in New Hampshire — looks good for Trump this time around. I would offer one caution, however. One reason that the surges for candidates like Rubio and Kasich looked so promising is because they were only occurring in the states about to vote. That made us more confident that they reflected voters seeking out new information about the candidates and weren’t just the result of changes in the news cycle. By contrast, while Trump is dominant in search traffic today, he’s less dominant in the 11 Republican Super Tuesday states. While he represents 73 percent of search traffic for the GOP candidates nationally, he has between 51 percent (Oklahoma) and 71 percent (Massachusetts) in the Super Tuesday states. Cruz and Rubio have about 50 percent more search traffic in Super Tuesday states than they do nationally, by contrast.

National 73% 10% 11% 3% 2%
Alabama 55 16 18 8 3
Alaska 68 13 10 6 4
Arkansas 54 18 20 6 3
Georgia 58 15 17 5 4
Massachusetts 71 8 11 3 7
Minnesota 64 11 18 4 3
Oklahoma 51 20 21 5 3
Tennessee 57 16 18 5 4
Texas 59 19 15 4 3
Vermont 70 9 9 5 7
Virginia 59 12 19 5 5
Trump’s search lead bigger nationally than in Super Tuesday states

Searches from 1:30 to 5:30 p.m. Eastern time

Source: Google Trends

5:52 PM
Super Tuesday Is A Turning Point

In this election cycle — if not in general — Super Tuesday represents a turn from the early contests. In Iowa and New Hampshire, the emphasis is on the interpretation of the election results. Rubio’s strong second-place finish in South Carolina, for example, garnered him a bunch of endorsements from Republican “elites.” The early contests award few delegates; their value is symbolic, based on informal but widely shared understandings of what each contest means for each candidate. Results acquire meaning because candidates meet, exceed or fail expectations, or because they show that candidates can compete for specific groups of voters. Trump pulled this off with evangelical voters in South Carolina; Sanders failed to do so among African-American voters there. In the early states, third place can be a victory, and a narrow first can be interpreted as defeat.

This begins to change, especially for Republicans, tonight. About half of the delegates necessary to win the nomination will be awarded in the 11 Super Tuesday contests. With a few exceptions, like Cruz’s performance in his home state, Texas, the big questions will probably be about delegate math rather than reading tea leaves about candidate appeal.

5:45 PM

And here’s how the states voting today shook out in the last competitive Democratic primary, in 2008:

Alabama Obama +14.4
American Samoa Clinton +14.7
Arkansas Clinton +43.8
Colorado Obama +34.3
Georgia Obama +53.3
Massachusetts Clinton +15.4
Minnesota Obama +34.2
Oklahoma Clinton +23.6
Tennessee Clinton +13.3
Texas Clinton +3.5
Vermont Obama +20.7
Virginia Obama +28.2
2008 Democratic primaries
5:40 PM

For your consideration, here’s how the Super Tuesday states voted in the Republican primaries in 2008 and 2012:

Alabama Huckabee +4.1
Alaska Romney +22.2
Arkansas Huckabee +40.2
Georgia Huckabee +2.3
Massachusetts Romney +24.9
Minnesota Romney +19.4
Oklahoma McCain +3.0
Tennessee Huckabee +2.7
Texas McCain +13.2
Vermont McCain +57.0
Virginia McCain +9.3
2008 Republican primaries
Alabama Santorum +5.2
Alaska Romney +3.2
Arkansas Romney +55.0
Georgia Gingrich +21.3
Massachusetts Romney +39.6
Minnesota Santorum +17.8
Oklahoma Santorum +5.8
Tennessee Santorum +9.1
Texas Romney +57.1
Vermont Romney +14.2
Virginia Romney +19.0
2012 Republican primaries
5:35 PM
The Evolution Of Super Tuesday

Super Tuesday still features contests in a bunch of Southern states, but it’s changed quite a bit from its origins. There are references to a big primary day as early as 1976, but the event as a coordinated regional primary for Southern states began in 1988. As Barbara Norrander describes in her book “Super Tuesday,” the idea behind Super Tuesday was that Southerners wanted more of a say in the Democratic Party’s nominees after Walter Mondale’s loss in the 1984 presidential election. They hoped that a coordinated primary might result in a more moderate — and perhaps Southern — nominee. That didn’t go so well; they ended up with Michael Dukakis that year.

As the parties have sorted into distinct ideological camps, however, and white Southern voters have fled the Democratic Party, the original motivation behind Super Tuesday is pretty much gone. This year, several of the Super Tuesday states have a Democratic electorate that’s more than 50 percent African-American. In a year when both Democratic candidates have been challenged about their records on race, a coordinated primary day that highlights the concerns of this constituency seems like a good idea. But it’s quite different from the original aim.

5:27 PM
Endorsements Starting To Help Trump In Our Model, For A Change

Trump leads in our forecasts in every state but Texas. But Rubio has been a little closer to Trump in what we call the “polls-plus” version of the model (as opposed to the “polls-only” version). That’s because the polls-plus version accounts for endorsements, where Rubio leads the GOP pack. This factor doesn’t hurt Trump as much as you might think, however, in part because Trump has started to pick up a few endorsements of his own. Furthermore, the model weights endorsements partly based on how recently they were made. When endorsements are weighted by recency category, Trump has actually moved up to second, behind Rubio but ahead of Cruz and Kasich.

Rubio 62% 60%
Trump 12 22
Cruz 14 10
Kasich 12 8
Carson <1 <1

It’s important to keep in mind that Trump’s share of endorsements is still unusually low given how well he’s done in the voting so far. But the factor is no longer the negative for him in our model that it once was.

5:25 PM

And here are the pre-Super Tuesday results for the Democrats:

2016 Iowa Democratic caucus results
2016 New Hampshire Democratic primary results
2016 South Carolina Democratic primary
2016 Nevada Democratic caucuses
5:20 PM

Before we start getting results tonight, refresh your memory — here’s what’s happened in the Republican race so far:

2016 Iowa Republican caucus results
2016 New Hampshire Republican primary results
2016 Nevada Republican caucuses
5:10 PM
How Many States Are Voting Today?

A:Republican voters in Colorado and Wyoming are going to the polls today. Does that mean those states are voting? It depends on who’s counting. Politico counts 661 Republican delegates at stake today, as do The New York Times and Time. Other outlets, including CNN, NPR, FiveThirtyEight and our sister site ABC News, count 595 delegates at stake. The 66-delegate difference is made up of 37 Colorado delegates and 29 from Wyoming.

The state Republican parties in both Colorado and Wyoming are holding caucuses today, but neither one directly, immediately determines how many delegates each candidate will get. Instead, Republicans in each state will vote today for delegates to conventions — at the county level in Wyoming, at the state level in Colorado. Those delegates may have stated preferences for a particular presidential candidate, but they are free to change their vote. So the Colorado and Wyoming delegates are known as unbound. The Republican parties in Colorado and Wyoming aren’t going to poll voters on their preferred candidate, so we won’t know who won the states.

Colorado’s Democrats are also caucusing today, but Wyoming’s Democrats don’t caucus until April 9. So if you see different counts of states voting today, Wyoming is probably the culprit.

There’s also some more confusion on the Democratic side. ABC and CNN say 1,015 Democratic delegates are up for grabs. Politico and NPR say 865, as do we. The difference here isn’t states, but the 150 superdelegates tied to the states voting today. Superdelegates are party insiders who might be influenced by the results today but can vote for whomever they like — they’re also known as unpledged delegates — so they aren’t really at stake.

Unfortunately, we’re not quite done tallying the confusion. The Times and the National Constitution Center say 1,034 Democratic delegates are at stake. That’s 19 more even than the counts that include superdelegates. The difference is mainly due to the vote that starts today — but ends a week from today — among Democratic U.S. citizens living abroad. That vote will determine 13 delegates. There are also eight superdelegates affiliated with the Global Presidential Primary — the eight members of the Democratic National Committee from Democrats Abroad. They’re really only semi-super: Their votes count for half as much as those of the other superdelegates.

So the Democrats Abroad constituency accounts for 17 of the 19-delegate gap. We’re still missing the other two. It turns out that discrepancy results from differences in counts of superdelegates in Colorado, American Samoa, Tennessee and Texas. The Constitution Center counts one fewer superdelegate from American Samoa than NPR does, while counting one more each from Colorado, Tennessee and Texas. What gives? Superdelegate counts are approximate. Superdelegates are people, generally officeholders or DNC members. Death, incapacitation or removal from office could throw off the counts. Don’t get too bothered about that tonight, because they’re unpledged anyway.

4:57 PM
Super Tuesday Viewers Guide

Welcome to FiveThirtyEight’s Super Tuesday live blog. We’ll be here all night — well, maybe not when they’re tallying caucus results in Alaska’s Unorganized Borough — but pretty late. Here are poll closing times (in EST), according to our colleagues at ABC News:

6 p.m. American Samoa Caucus 6
7 p.m. Georgia Primary 102 76
Vermont Primary 16 16
Virginia Primary 95 49
8 p.m. Alabama Primary 53 50
Massachusetts Primary 91 42
Oklahoma Primary 38 43
Tennessee Primary 67 58
8:30 p.m. Arkansas Primary 32 40
9 p.m. Colorado Caucus 66
Minnesota Caucus 77 38
Texas Primary 222 155
Midnight Alaska Caucus 28
Super Tuesday poll closing times

Source: ABC News

We expect Donald Trump and Hillary Clinton to have great nights, but there’s something to watch at almost every hour.

Virginia, where polls close at 7 p.m. Eastern time, once looked like a potential opportunity for Marco Rubio, although he’s slipped further behind Trump in the polls there. And Georgia offers the second-biggest delegate prize of the night, after Texas.

Oklahoma, where polls close at 8 p.m., is our pick for most underrated state of the night; the race between Clinton and Bernie Sanders is a true toss-up, according to our forecast, while polls show both Rubio and Ted Cruz a little closer to Trump in Oklahoma than they are in other states. The other state to watch is Massachusetts, which is one of the states Sanders probably needs to win to be on track for the nomination, but he trails Clinton in recent polls there.

The 9 p.m. poll closings represent Trump’s best chance to actually lose states. He trails Cruz in Texas polls, although some recent surveys have shown the gap closing. And as a Midwestern caucus, Minnesota could offer an opportunity for either Cruz or Rubio, as Iowa did a month ago.

Lastly, our usual reminder that polling can be a rough enterprise in the primaries, especially in the multi-candidate race that the GOP now has. But that error could work in either direction. It could be that Trump and Clinton have even better nights than we’re expecting or that Sanders, Rubio and Cruz pull off some upsets where polls deemed them to be unlikely.

Stick with us all night!

Tweet us a question or comment